Reproducibility of a food frequency questionnaire used in the New York University Women's Health Study: effect of self-selection by study subjects.

OBJECTIVE: The aim of the study was to investigate the reproducibility of a food frequency questionnaire (FFQ) used in a cohort of 14,290 women enrolled in the NYU Women's Health Study. DESIGN: A subset of 474 cohort subjects who completed the dietary questionnaire at baseline (FFQ-1) were approached again on the occasion of a second visit to the mammography study centre and asked to complete the questionnaire a second time (FFQ-2) two to four years after FFQ-1. Two to three months later the questionnaire was mailed to the subjects, and they were asked to complete it a third time (FFQ-3). SETTING: A breast cancer screening clinic. SUBJECTS: Of the 474 subjects selected, 100% completed the second questionnaire while only 56% completed and mailed back FFQ-3. This made it possible to compare the long-term reproducibility of dietary intake measurements and baseline dietary habits between the two groups of subjects 'respondents', who agreed to complete the questionnaire a third time, and 'non-respondents', who did not. RESULTS: Among respondents (56% of study subjects), energy-adjusted correlation coefficients for short-term reproducibility between FFQ-2 and FFQ-3 ranged from 0.50-0.64 for nutrients, and from 0.44-0.67 for foods. The long-term reproducibility was lower, ranging from 0.36-0.53 for nutrients, and from 0.31-0.48 for specific food groups. Among those who did not respond to FFQ-3, crude correlations for long-term reproducibility, unadjusted for energy intake, were generally lower than among respondents. Nevertheless, after adjustment for energy intake, correlations for long-term reproducibility (FFQ-2 to FFQ-1) were of similar magnitude in both groups. In addition, 'non-respondents' reported lower intake of fruit and vegetables and higher intake of meat. CONCLUSIONS: The results of this study suggest that subjects who volunteer to participate in substudies on the validity or reproducibility of dietary questionnaire measurements may tend to provide more accurate responses to the questionnaire. The phenomenon seems related more to accuracy of reporting of absolute intake levels than of the relative composition of diet, self-selection may be associated with differences in dietary habits.

Relation of serum levels of testosterone and dehydroepiandrosterone sulfate to risk of breast cancer in postmenopausal women.

The authors examined the relation between postmenopausal serum levels of testosterone and dehydroepiandrosterone sulfate (DHEAS) and subsequent risk of breast cancer in a case-control study nested within the New York University Women's Health Study cohort. A specific objective of their analysis was to examine whether androgens had an effect on breast cancer risk independent of their effect on the biologic availability of estrogen. A total of 130 cases of breast cancer were diagnosed prior to 1991 in a cohort of 7,054 postmenopausal women who had donated blood and completed questionnaires at a breast cancer screening clinic in New York City between 1985 and 1991. For each case, two controls were selected, matching the case on age at blood donation and length of storage of serum specimens. Biochemical analyses were performed on sera that had been stored at -80 degrees C since sampling. The present report includes a subset of 85 matched sets, for whom at least 6 months had elapsed between blood donation and diagnosis of the case. In univariate analysis, testosterone was positively associated with breast cancer risk (odds ratio (OR) for the highest quartile = 2.7, 95% confidence interval (CI) 1.1-6.8, p < 0.05, test for trend). However, after including % estradiol bound to sex hormone-binding globulin (SHBG) and total estradiol in the statistical model, the odds ratios associated with higher levels of testosterone were considerably reduced, and there was no longer a significant trend (OR for the highest quartile = 1.2, 95% CI 0.4-3.5). Conversely, breast cancer risk remained positively associated with total estradiol levels (OR for the highest quartile = 2.9, 95% CI 1.0-8.3) and negatively associated with % estradiol bound to SHBG (OR for the highest quartile = 0.05, 95% CI 0.01-0.19) after adjustment for serum testosterone levels. These results are consistent with the hypothesis that testosterone has an indirect effect on breast cancer risk, via its influence on the amount of bioavailable estrogen. No evidence was found of an association between DHEAS and risk of breast cancer in postmenopausal women.

Endogenous estrogens and risk of breast cancer by estrogen receptor status: a prospective study in postmenopausal women.

A positive association between postmenopausal serum levels of total estradiol, percentage of free estradiol, and percentage of estradiol not bound to sex hormone-binding globulin (SHBG) and breast cancer risk was recently reported by the New York University Women's Health Study (P. Toniolo et al., J. Natl. Cancer Inst., 87: 190-197, 1995). Data from this prospective study are used to assess whether the observed associations differ according to estrogen receptor (ER) status of the tumor. Between 1985 and 1991, 7063 postmenopausal women donated blood and completed questionnaires at a large breast cancer screening clinic in New York City. Before 1991, 130 cases of first primary breast cancer were identified by active follow-up of the cohort. For each case, two controls were selected, matching the case on age at first blood donation and length of storage of specimens. Biochemical analyses were performed on sera that had been stored at -80 degrees since sampling. ER information was abstracted from pathology reports. Separate statistical analyses were conducted of ER-positive, ER-negative, and ER-unknown groups (53, 23, and 54 matched sets, respectively). In each of the 3 groups, the mean estradiol and the mean percentage of free estradiol were greater (21-28% and 6-7%, respectively) in cases than in controls. Conversely, the mean percentage of estradiol bound to SHBG was 9-12% lower in cases than in controls. The logistic regression coefficients measuring the strength of the association between estradiol and its free and SHBG-bound fractions and breast cancer risk were similar in the ER-positive, ER-negative, and ER-unknown groups. These data suggest that in postmenopausal women, the association of endogenous estrogens with breast cancer risk is independent of the ER status of the tumor. This result is more compatible with the hypothesis of a progression from ER-positive to ER negative tumors than with the hypothesis that ER status identifies two distinct types of breast cancer.

Height and mortality after myocardial infarction.

A case-control analysis was conducted to determine the relationship between height and mortality among patients enrolled in the already completed Beta Blocker Heart Attack Trial (BHAT). In a basic model including height (continuous) and relevant covariates the relative risk (RR) per 4-inch reduction in height (approximately 1 standard deviation) was 1.18 (95% confidence interval, 0.92 to 1.51). When sex was considered, the effect of short stature on mortality was found to be restricted to male subjects. The male RR per 4-inch reduction in height was 1.26 (0.96 to 1.63) whereas for women it was 0.89 (0.49 to 1.59). In males not randomized to propranolol (untreated) the effect was further modified with a RR per 4-inch reduction in height of 1.41 (1.00 to 1.99). It is hypothesized that short stature could be a marker for factors operating as far back as childhood that predispose males to mortality from coronary heart disease in later life.

Estimating the reliability of an exposure variable in the presence of confounders.

In this paper we discuss estimation of the reliability of an exposure variable in the presence of confounders measured without error. We give an explicit formula that shows how the exposure becomes less reliable as the degree of correlation between the exposure and confounders increases. We also discuss biases in the corresponding logistic regression estimates and methods for correction. Data from a matched case-control study of hormone levels and risk of breast cancer are used to illustrate the methods.

Epidemiologic follow-up studies of breast augmentation patients.

Although hundreds of thousands of women in this country have had augmentation mammaplasty, little is known about long-term effects. Clinical studies have documented a decreased ability to detect breast lesions in women with implants, leading to concerns regarding breast cancer risk. There is also anecdotal evidence that implants might have effects on a variety of immune conditions. More recently, concern over carcinogenic effects has heightened given findings that the polyurethane foam, used in a minority of implants to envelope silicone gel, contains chemicals linked to cancer in laboratory animals. Only a few epidemiologic studies on long-term effects have been published, and all have had methodologic limitations, including the possibility of inappropriate comparison rates, limited and/or incomplete follow-up, absence of information on patients characteristics, and lack of specific information on types of implanted material. Several case-control and cohort studies are currently underway which are attempting to overcome methodologic limitations of previous studies. Past descriptive and analytic studies are reviewed. In addition, ongoing follow-up efforts are discussed, with attention given to the methodologic adjuncts necessary for allowing valid assessments of long-term disease effects.

A prospective study of endogenous estrogens and breast cancer in postmenopausal women.

BACKGROUND: Circumstantial evidence links endogenous estrogens to increased risk of breast cancer in women, but direct epidemiologic support is limited. In particular, only a few small prospective studies have addressed this issue. PURPOSE: Our purpose was to assess breast cancer risk in relation to circulating levels of the two major endogenous estrogens, estrone and estradiol, measured before the clinical onset of the disease. METHODS: The association between serum levels of estrogens and the risk of breast cancer was examined in a prospective cohort study of 14,291 New York City women, 35-65 years of age, who received screening for breast cancer at the time of blood sampling and who had not been diagnosed with breast cancer. During the first 5 1/2 years of study, we identified 130 breast cancers among the postmenopausal group (7063 women, 35,509 person-years). The case subjects and twice as many postmenopausal control subjects were included in a case-control study nested within the cohort. Biochemical analyses for percent free estradiol, percent estradiol bound to sex hormone-binding globulin (SHBG), total estradiol, estrone, and follicle-stimulating hormone were performed on sera that had been kept at -80 degrees C since sampling. RESULTS: For increasing quartiles of total estradiol, the odds ratio (ORs) of breast cancer, as adjusted for Quetelet index (weight in kilograms divided by the square of the height in meters), were 1.0, 0.9, 1.8, and 1.8 (P value for trend = .06); the ORs for increasing quartiles of estrone were 1.0, 2.2, 3.7, and 2.5 (P value for trend = .06). For increasing quartiles of free estradiol, defined as the fraction of estradiol that is not bound to proteins, the Quetelet index-adjusted ORs of breast cancer were 1.0, 1.4, 3.0, and 2.9 (P value for trend < .01). When we considered the percent of estradiol bound to SHBG, the Quetelet index-adjusted ORs were 1.0, 0.70, 0.40, and 0.32 (P value for trend < .01), thus suggesting a strong protective effect. These associations persisted or became even stronger when analyses were restricted to women whose samples had been drawn 2 or more years before breast cancer diagnosis. CONCLUSIONS: These data represent the first confirmation in a large prospective epidemiologic study of a link between circulating estrogens and breast cancer risk. Although estrogen levels appeared to fall within the conventional limits of normality in all women under study, those who subsequently developed breast cancer tended to show higher levels of estrone, total estradiol, and free estradiol, and a lower percent of estradiol bound to SHBG than women who remained free of cancer. IMPLICATIONS: Factors that increase endogenous estrogen production or reduce the binding of estradiol to SHBG may increase a woman's risk of developing breast cancer later in life.

The antinociceptive effect of S-(+)-ibuprofen in rabbits: epidural versus intravenous administration.

This study was designed to determine whether systemic absorption plays any role in the antinociceptive effect of epidural (EP) sodium S(+)-ibuprofen (IB). One week after surgical implantation of EP catheters, six rabbits were given EP injections with either normal saline (NS) 0.4 mL or IB 10 mg in 0.4 mL NS (Group 1) on separate days. Each animal was injected with IB 10 mg intravenously (i.v.) on another day. Six control rabbits (Group 2) had neither surgery nor any injection. Analgesic testing was performed using electric stimulation through two electrocardiogram (ECG) skin electrodes with built-in adhesive, attached to shaved hip areas using 50 V, 1 Hz, 3 ms, before and 0.5,1,2 and 3 h after injection in Group 1, and in similar times in controls. The 95% confidence intervals (CI) of the mean difference between baseline and maximal nociceptive response latency of all groups were compared using analysis of covariance (ANCOVA) adjusted for baseline measurements. This comparison covered all possible pairs among all groups. Significant antinociceptive effects were seen after EP IB but not after control or i.v. IB. Neither motor dysfunction nor evidence of systemic toxicity or neurotoxicity was observed in any animal.

Premenopausal estradiol levels and the risk of breast cancer: a new method of controlling for day of the menstrual cycle.

Levels of total estradiol in premenopausal women vary widely over the course of the menstrual cycle with a spike at the time of ovulation and dissimilar patterns pre- and post-ovulation. Evaluating the association between breast cancer and premenopausal measurements of total estradiol when the measurements cannot be taken on a uniform day of the cycle is therefore a difficult methodological challenge. In a matched case-control study of breast cancer nested within a prospective study, premenopausal serum samples obtained up to 7 years before breast cancer diagnosis were available for total estradiol assay. By fitting a three-piece spline model that regressed the logarithm of total estradiol (ln estradiol) on day of menstrual cycle, the authors were able to adjust the measurements for day of the cycle on which they were collected by expressing them in terms of the number of standard deviations above or below the fitted ln estradiol value for that day. Applying the adjusted measurements to the nested case-control study, they found evidence of a 1.5 to 2-fold risk for women in the upper two tertiles of ln estradiol relative to women in the lowest tertile. Conditional logistic regression analysis for day-of-cycle-adjusted ln estradiol treated as a continuous variable resulted in a relative risk estimate of 1.19 (95% confidence interval 0.91-1.55) per standard-deviation increase in adjusted ln estradiol.

Consumption of meat, animal products, protein, and fat and risk of breast cancer: a prospective cohort study in New York.

Epidemiologic studies have focused on the association between diet and breast cancer with conflicting results. Whereas a majority of case-control studies indicate a role for the intake of total fat and saturated fat, most prospective cohort studies either are negative or indicate very modest associations. Only a few authors have examined the role of meat intake in relation to breast cancer risk. The aim of this study was to examine the relation between risk of breast cancer and dietary intake of meat, animal products, fat, and protein. Between 1985 and 1991, we recruited 14,291 New York City women in a prospective cohort study of endogenous hormones, diet, and cancer in which they reported on their recent diet using a food frequency questionnaire self-administered at enrollment. From the cohort, 180 invasive breast cancer cases diagnosed before December 1990 and five times as many controls, individually matched by age, calendar time at enrollment, menopausal status, and, if premenopausal, phase of menstrual cycle, were included in a nested case-control study. There was an evident increase in the relative risk (RR) of breast cancer for increasing consumption of meat. Women in the upper quintile of meat consumption, as compared with the lowest quintile, had an energy-adjusted RR of 1.87 (95% confidence interval = 1.09-3.21). There was a modest RR increase in the upper quintile of total and saturated fat and no apparent association for other types of fat, protein, dairy products, poultry, or fish.(ABSTRACT TRUNCATED AT 250 WORDS)

Reliability of measurements of total, protein-bound, and unbound estradiol in serum.

Estradiol (E2) circulates in the blood in three states: unbound (U-E2), bound to sex-hormone binding globulin (SHBG-E2), and bound to albumin. There is evidence to support the concept that only U-E2 and albumin-bound E2, are bioavailable (i.e., rapidly extracted by tissues). A case-control study nested within a large cohort of women, in which we are examining the effect of estrogens on breast cancer risk, offered the opportunity to assess the reliability of measurements of E2, the percentage of SHBG-E2, and the percentage of U-E2 based on multiple annual serum specimens. Long-term (1-2 year) reliability, as estimated by the intraclass correlation coefficient, was assessed in a subgroup of 71 premenopausal and 77 postmenopausal controls for whom two or three serum specimens were assayed. In postmenopausal women the intraclass correlation coefficient for a single measurement of total E2 was only 0.51. As for the percentage of SHBG-E2, intraclass correlation coefficients were 0.83 and 0.94, and for U-E2, 0.72 and 0.77 in the premenopausal and postmenopausal groups, respectively. These data suggest that, whereas single determinations of total E2 are insufficient to reliably estimate a woman's true mean level, a single measurement of the percentage of SHBG-E2 or U-E2 is adequate to assess bioavailability of E2 in an epidemiological study, irrespective of day of the menstrual cycle.

The influence of medical conditions associated with hormones on the risk of breast cancer.

Medical conditions related to hormonal abnormalities were investigated in a case-control study of breast cancer among women who attended a screening centre. Information was obtained by telephone interview regarding physician-diagnosed medical conditions such as thyroid or liver diseases, diabetes, and hypertension, as well as hirsutism, acne, galactorrhoea, and reproductive, menstrual, and gynaecological factors. Results are presented for 354 cases and 747 controls. Women with fertility problems who never succeeded in becoming pregnant were at significantly increased breast cancer risk (adjusted odds ratio [OR] = 3.5; 95% confidence interval [CI]:1.1-10.9). An elevated cancer risk was also associated with having excess body hair (OR = 1.5; 95% CI:1.0-2.3), or having excess body hair in addition to persistent adult acne (OR = 6.8; 95% CI:1.7-27.1). Recurrent amenorrhea (OR = 3.5; 95% CI:1.1-11.5), and a treated hyperthyroid condition (OR = 2.2; 95% CI:1.1-4.4) were significantly associated with risk. A non-significant elevation of risk was present for endometrial hyperplasia (OR = 1.8; 95% CI: 0.8-4.0). There was a suggestion of an association between a history of galactorrhoea and breast cancer risk (OR = 2.0; 95% CI:0.8-4.9) among premenopausal women. No associations were found with other medical or gynaecological factors. The possibility that some of these findings are due to chance cannot be excluded because of the problem of multiple comparisons.

Reliability of serum prolactin measurements in women.

Prolactin, a hormone indispensable for milk secretion, has been shown to enhance the development and growth of mammary tumors in rodents; however, its importance in human breast cancer is uncertain. Serum prolactin levels are known to fluctuate considerably under normal conditions, and lack of precision in the hormone measurements may have contributed to the largely negative findings in humans to date. The purpose of this study was to investigate the reliability of prolactin measurements in women using stored serum from an ongoing prospective study of breast cancer. Separate groups of postmenopausal and premenopausal women who donated multiple blood samples at approximately 1-year intervals were studied. The reliability of a single log prolactin determination, as measured by the intraclass correlation coefficient, was 0.76 for the postmenopausal women (95% confidence interval, 0.66-0.85) and 0.48 for the premenopausal women (95% confidence interval, 0.31-0.62). These findings suggest that a single measurement is sufficient to characterize the serum prolactin level of postmenopausal women for epidemiological research. For premenopausal women, however, multiple samples are desirable. Controlling for phase of the menstrual cycle does not appear to substantially improve the reliability of premenopausal measurements.

Sources of variability in waist and hip measurements in middle-aged women.

The reliability of single measurements of waist and hip circumference and the ratio of waist circumference to hip circumference (a widely used measure of body fat distribution) has not been fully examined. The authors analyzed measurements of waist and hip circumference, as well as self-reported weight and height, repeated 3-6 times between 1986 and 1991 among 1,851 participants in the New York University Women's Health Study. Quetelet index (weight (kg)/height (m)2) was positively correlated with waist circumference (r = 0.88), hip circumference (r = 0.89), and waist/hip ratio (r = 0.52). Mean weight was positively correlated with the within-subject variance of waist circumference (r = 0.27) and, to a lesser degree, with the within-subject variance of hip circumference (r = 0.08) and waist/hip ratio (r = 0.10). The within-subject variance of weight was positively correlated with the within-subject variance of waist (r = 0.30) and hip (r = 0.23) measurements, and less so with waist/hip ratio (r = 0.05). Intraclass correlations for waist, hip, and waist/hip ratio were 0.89, 0.81, and 0.74, respectively; adjustment for Quetelet index reduced the intraclass correlations for waist and hip measures by 33% and 48%, respectively. Such adjustment can provide a more realistic determination of the reliability associated with an exposure variable in the design and analysis of studies investigating the relation between body fat distribution and disease.

Intermittent ST depression and mortality after myocardial infarction.

BACKGROUND: We conducted a case-control analysis to determine the contribution made to mortality by intermittent ST depression (STD) among patients enrolled in the already completed Beta Blocker Heart Attack Trial. METHODS AND RESULTS: STD was determined by computer analysis of 24-hour ECG tapes as a change in ST level by +/- 0.1 mV or more from the median value of ST of all normally conducted complexes for greater than or equal to 1 minute. All computer-detected ST events were verified by trained readers. To estimate risk of dying associated with STD, 261 deaths were compared with controls matched for age, sex, drug status, and time elapsed since acute myocardial infarction. In a model including relevant covariates, STD had a relative risk (RR) of 1.73 (95% confidence interval, 1.09-2.73). The RR was 2.56 (1.39-4.71) in untreated patients and 0.98 (0.48-2.00) in propranolol-treated patients. A history of angina, although not independently significant, was found to enhance these RRs. A gradient of risk was shown in the placebo group by a RR of 1.91 in those with 1-30 minutes of STD and 4.33 in those with greater than 30 of STD (p = 0.001, trend test). CONCLUSIONS: The findings in this large study show a significant contribution to mortality among untreated early post-myocardial infarction survivors from transient STD on 24-hour monitoring. The absence or reduction of effect in the treated group also suggests an anti-ischemic mechanism by which propranolol exerts a protective effect on mortality. Trials to determine whether reduction of STD improves survival would be warranted.

Use of human data in quantitative risk assessment of carcinogens: impact on epidemiologic practice and the regulatory process.

Epidemiologic data are increasingly being used to assess cancer risk from chemicals as their value is recognized and as more and better studies become available. Weight-of-evidence approaches are now available for classifying the experimental and epidemiological evidence regarding human carcinogenicity. When the human data are extensive and of good quality, they should be given substantial weight in assessing risk. Both the positive and the negative epidemiologic data should be used in a quantitative risk assessment (QRA), because only then can an unbiased risk assessment be derived. Good-quality epidemiological studies are those with sound methodology, lack of bias, long enough follow-up times to observe a carcinogenic response, adequate exposure information, and dose-response information. Before a lack of carcinogenicity can be inferred, it is essential that the exposures be of substantial duration and intensity, and that the number of exposed persons be reasonably large. Epidemiologists need to give more attention to exposure assessment, because lack of quantitative exposure information is often the limiting factor that prevents the use of epidemiologic data in QRA. Development of methods to estimate historical workplace exposure intensities from surrogate industrial hygiene variables should receive high research priority, since they have the potential to increase the usefulness in QRA of many epidemiologic studies that have limited exposure information. Several frequently used surrogates for exposure measurements have limitations or pitfalls in their use. In particular, the use of "ever/never" exposed has a large potential to produce falsely negative results by means of a "dilution" effect, especially in the common case where the exposure distribution is skewed. Duration of exposure (rather than duration x intensity) may also give misleading results. There is little information to suggest that synergistic exposures to multiple toxicants in an industrial environment are likely to invalidate QRAs, probably because few studies have identified a group of workers with major workplace exposures to multiple carcinogens that cause the same type of cancer. Most of the interactive effects which have been identified to date are between smoking and some occupational carcinogen, so this possibility needs careful evaluation for smoking-related diseases. It is important to evaluate dose-response gradients in a QRA to obtain maximum precision and accuracy in the resulting risk coefficient. The analysis should take into account an appropriate cancer induction period. Various methods to account for cancer induction times are compared; those that incorporate a lag period or model the induction-time distribution are superior to other methods.(ABSTRACT TRUNCATED AT 400 WORDS)

Expression of the CYP1A1 gene in peripheral lymphocytes as a marker of exposure to creosote in railroad workers.

We have conducted a pilot study to assess levels of cytochrome CYP1A1 gene expression in human peripheral lymphocytes as a molecular biomarker assay for polycyclic hydrocarbon exposure. Basal and 3-methylcholanthrene-induced levels of gene expression were measured by standard slot-blot mRNA analyses in mitogen-stimulated cultures of peripheral blood lymphocytes from creosote-exposed railroad workers and unexposed control subjects. Dermal and inhalation exposure of workers to creosote may vary substantially as a function of working conditions related to temperature. Therefore, blood specimens were collected from separate groups during the winter, fall, and summer. Basal and induced CYP1A1 gene expression levels were not elevated in workers from any of the three seasonal studies. However, induced/basal (inducibility) CYP1A1 mRNA ratios from workers sampled in the summer (when actual exposures were greatest) were significantly higher when compared to those of controls (P < 0.01). These studies demonstrate the potential usefulness of specific gene expression assays in human peripheral lymphocytes for the assessment of carcinogen exposure in human populations.

Hair dye use and breast cancer: a case-control study among screening participants.

To investigate whether hair dye use increases the risk of breast cancer, a case-control study was conducted among patients attending a screening center in New York City. The study group consisted of 398 breast cancer cases identified at the screening center between 1977 and 1981, and 790 randomly selected controls screened during the same period. Subjects were interviewed by telephone to obtain information on known risk factors for breast cancer, along with a complete history of hair dye use detailing type of dye, color, duration, frequency, and temporal periods of use. Most subjects (77%) had used hair dye at least once, 38% of the subjects at least 100 times. However, little increased risk of breast cancer was found among hair dye users. The adjusted odds ratio for ever having used hair dye was 0.8 (95% confidence interval 0.6-1.1), and there was no evidence of a trend in risk with increasing number of hair dye uses. The results were the same whether all past exposures were considered or only exposures more than 10 years before disease. Breast cancer risk did not increase with increasing intensity of exposure, as measured by frequency of use or darkness of color. No effect was seen for different types or colors of dye, or for use during different periods of reproductive life. Although personal hair dye use was unrelated to breast cancer risk, there was an adjusted odds ratio of 3.0 (95% confidence interval 1.1-7.8) for 5 or more years of work as a beautician. Overall, the results of this study, taken in conjunction with the findings of other epidemiologic studies, do not implicate hair dye use as an important cause of human breast cancer.